Printing buglet for latin1 encoded strings
x <- c("ao?t", "?l?ve", "?a va")
y <- iconv(x, "UTF-8", "latin1")
x
#> [1] "ao?t" "?l?ve" "?a va"
y
#> [1] "ao?t" "?l?ve" "?a va"
(note the difference in spacing)
On both R-release (mac) and R-devel (linux). nchar() and format() both look ok.
Hadley